Echocardiographic HFpEF phenotype in people with HIV: prevalence and associated factors in the COCOMO cohort

Background

People with HIV (PWH) are at increased risk of heart failure despite effective antiretroviral therapy. Heart failure with preserved ejection fraction (HFpEF) is increasingly recognized in aging populations; however, its prevalence and associated clinical features in contemporary PWH are not well defined.

Purpose

To assess the prevalence and correlates of an echocardiographic HFpEF phenotype in PWH using a modified Heart Failure Association (HFA)-PEFF score.

Methods

We included 528 PWH with LVEF ≥50% from a prospective, observational cohort. All participants underwent comprehensive echocardiography, physical examination, blood sampling, and questionnaire-based assessment of dyspnea. The modified HFA-PEFF score was calculated using only the structural and functional echocardiographic domains (excluding the natriuretic peptide domain), with a maximum of 2 points per domain. Participants were classified as low (0–1 points), intermediate (2–3 points), or high (4 points) probability of HFpEF. Dyspnea was categorized as mild (hurrying/uphill) or moderate/severe (walking at usual pace or at rest).

Unadjusted trends in baseline characteristics across modified HFA-PEFF categories were assessed using regression models treating score category as an ordinal variable. Associations between clinical and HIV-related characteristics and an echocardiographic HFpEF phenotype (HFA-PEFF score 2–4 vs 0–1) were evaluated using multivariable logistic regression adjusted for age, sex, BMI, hypertension, diabetes, eGFR, and smoking status.

Results

Mean age was 53 ± 11 years, and 88% were male. Modified HFA-PEFF scores were low in 287 (54%), intermediate in 221 (42%), and high in 20 (4%). Increasing modified HFA-PEFF score was associated with older age, higher BMI, hypertension, diabetes, lower eGFR, longer HIV duration, prior AIDS, and higher NT-proBNP and troponin T, whereas dyspnea prevalence did not differ across categories (Table 1).

In multivariable logistic models, age (aOR 2.39 [1.88–3.07] per 10-year increase), BMI (aOR 1.52 [1.11–2.09] per 5 kg/m² increase), hypertension (aOR 2.04 [1.35–3.10]) and prior AIDS (aOR 2.19 [1.22–4.00]) were associated with an HFpEF phenotype. HIV duration, CD4 nadir <200 cells/μL, biomarkers, and dyspnea were not associated (Figure 1).

Conclusion

In PWH, a modified HFA-PEFF approach identified an HFpEF echocardiographic phenotype in nearly half of participants despite a predominantly middle-aged cohort. The phenotype was independently associated with cardiometabolic comorbidity and prior AIDS, but not with symptoms or biomarkers.